Real-world
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Motion Magnification | Real-world (test) | Motion Error0.26 | 78 | |
| Prototype Selection | Real-world datasets | Average Rank1.93 | 54 | |
| Anomaly Detection | Real-world datasets | AUCROC0.921 | 33 | |
| Peg Insertion | Real-world | Success Rate91.7 | 25 | |
| Articulated Joint Estimation | Real-world dataset | mAP (5° 5cm)92.23 | 24 | |
| Part Segmentation | Real-world dataset | mIoU (75%)98.79 | 24 | |
| ProCams simulation | Real-world Novel Viewpoints | PSNR26.1 | 16 | |
| ProCams simulation | Real-world benchmark dataset Trained Viewpoints | PSNR26.65 | 16 | |
| Pick & Place | Real World | Success Rate94.12 | 15 | |
| RMFS Order Allocation and Robot Scheduling | Real-World Medium (test) | Objective Value653.41 | 14 | |
| Surface Normal Estimation | Real-world Average | MAE (°)12.54 | 14 | |
| Semantic Segmentation | Real-world | Dice0.5955 | 14 | |
| Anomaly Detection | Real-world | Average Time Cost (s)1.1 | 13 | |
| Hang Cups | Real-World Unseen | Success Rate88 | 13 | |
| Mosaiced and PAN image fusion | Real-world | QNR0.8709 | 11 | |
| Open drawer | Real-world (test) | Success Rate96 | 11 | |
| Insertion | Real-World | Success Rate85 | 11 | |
| Modular Robot Path Planning and Execution | Real-world (val) | Simulation Cycle Time (s)1.27 | 10 | |
| pick and place bear | Real-world | Success Rate0.9 | 10 | |
| Video Frame Reconstruction | Real-World | FID184.2509 | 10 | |
| Event-based video frame reconstruction | Real-World | MSE0.0612 | 10 | |
| Passive mapping | Real world (test) | F1 Score55.5 | 10 | |
| Imbalance Ratio Preservation | real-world datasets source (train) | Mean |ΔIR|1.303 | 9 | |
| Average Manipulation Performance | Real-world | Average Success Rate78.2 | 9 | |
| Pouring | Real-world | Success Rate73.5 | 9 |
Showing 25 of 403 rows
...